diff options
author | Christian Pointner <equinox@helsinki.at> | 2015-07-16 15:45:26 (GMT) |
---|---|---|
committer | Christian Pointner <equinox@helsinki.at> | 2015-07-16 15:45:26 (GMT) |
commit | fc91f05b685bc480845b77b273b5072a1ef54581 (patch) | |
tree | 69a09bcc9734f28f47df54d560e16ec926e01ba6 /rhimport-tr | |
parent | 018341186bfce227c6102d12b99878c1d3885537 (diff) |
added importer for tierrecht
Diffstat (limited to 'rhimport-tr')
-rwxr-xr-x | rhimport-tr | 189 |
1 files changed, 189 insertions, 0 deletions
diff --git a/rhimport-tr b/rhimport-tr new file mode 100755 index 0000000..d86bdb9 --- /dev/null +++ b/rhimport-tr @@ -0,0 +1,189 @@ +#!/usr/bin/perl -w +# +# +# rhimport +# +# Copyright (C) 2009 Christian Pointner <equinox@helsinki.at> +# +# This file is part of rhimport. +# +# rhimport is free software: you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation, either version 3 of the License, or +# any later version. +# +# rhimport is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with rhimport. If not, see <http://www.gnu.org/licenses/>. +# + +use strict; +use DBI; +use Date::Calc; +use XML::Feed; +use XML::Feed::Entry; +use XML::Feed::Content; +use XML::Feed::Enclosure; +use File::Fetch; +use IO::Handle; +use IPC::Open3; +use HTML::Entities; + +use lib '/usr/local/share/rhimport/'; +use rhimport; + +my $DBHOST = "airplay"; +my $DBUSER = "rivendell"; +my $DBPW = "lldriven"; +my $DB = "rivendell"; + +my $STAT_FILE = $ENV{'HOME'} . "/rhimport-tr.stat"; +my $RSS_URL = "http://cba.fro.at/series/tierrechtsradio/feed?c=Kfs2IoV2Wmd"; +$XML::Feed::MULTIPLE_ENCLOSURES=1; +my $PV_ID = '506'; + +binmode(STDIN, ":utf8"); +binmode(STDOUT, ":utf8"); +binmode(STDERR, ":utf8"); + +if($#ARGV >= 0 && $ARGV[0] eq 'last') { + print "!!!This is the last attempt, there won't be a retry on error!!!\n" +} + +my $user = `/usr/bin/id -un`; +$user =~ s/\n//; +my $group = "tierrecht"; + +my $dbh = DBI->connect("DBI:mysql:$DB:$DBHOST","$DBUSER","$DBPW") or die "Database Error: $DBI::errstr"; +my @allowed_dbs = rhimport::get_dropboxes($dbh, $user, $group); + +my $idx = 0; +if(scalar(@allowed_dbs) != 1) { + print "found more or less than 1 Dropbox for this group?!\n"; + $dbh->disconnect(); + exit 1; +} +my $dropbox = $allowed_dbs[$idx]->{'PATH'}; +my $to_cart = $allowed_dbs[$idx]->{'TO_CART'}; + +my @today = Date::Calc::Standard_to_Business(Date::Calc::Today()); +if($today[2] == 5) { + print "please don't run this script on fridays!\n"; + $dbh->disconnect(); + exit 1; +} +my @broadcast_day = Date::Calc::Business_to_Standard($today[0], $today[1], 5); +if($today[2] < 5) { + @broadcast_day = Date::Calc::Add_Delta_Days($broadcast_day[0], $broadcast_day[1], $broadcast_day[2], -7); +} +my @import_date = Date::Calc::Add_Delta_Days($broadcast_day[0], $broadcast_day[1], $broadcast_day[2], 7); + +print "day of original broadcast: " . Date::Calc::Date_to_Text(@broadcast_day) . "\n"; +print "day of Radio Helsinki broadcast: " . Date::Calc::Date_to_Text(@import_date) . "\n"; + +my $bd = sprintf("%04d-%02d-%02d", $broadcast_day[0], $broadcast_day[1], $broadcast_day[2]); +my $bdfile = sprintf("^%04d%02d%02d", $broadcast_day[0], $broadcast_day[1], $broadcast_day[2]); + +print "looking for files from $bd in RSS Feed\n"; +print " -> $RSS_URL\n"; + +my $feed = XML::Feed->parse(URI->new($RSS_URL)) + or die "Error fetching feed: " . XML::Feed->errstr; + +my $file = ""; +my $out_file = ""; +my $i = 0; +for my $entry ($feed->entries) { + $i++; + next unless $entry->enclosure; + my $j = 0; + for my $enclosure($entry->enclosure) { + $j++; + if($enclosure->type eq "audio/mpeg" || $enclosure->type eq "audio/ogg") { + print "$i/$j: (" . $enclosure->type . ", " . $enclosure->length . ") " . $enclosure->url . "\n"; + my $url = $enclosure->url; + if($url =~ /^(.*)_cut(\.[^\.]+)$/) { + $url = $1 . $2; + } + + my $ff = File::Fetch->new(uri => $url); + next unless (uc($ff->output_file) =~ $bdfile); # file not from correct day + + my $sum_title = decode_entities($entry->title); + my $sum_text = decode_entities($entry->content->body); + + my $current_file = `cat $STAT_FILE`; + if($current_file eq $ff->output_file) { + print "Already downloaded file of today\n"; + $dbh->disconnect(); + exit 0; + } + $out_file = $ff->output_file; + if(!rhimport::check_file_extension($out_file)) { + print "\n\nThe extension of the matching file '". $out_file . "' seems to be wrong - manual import necessary!!!\n"; + $dbh->disconnect(); + exit 1; + } + + print " --> " . $bdfile . ": downloading " . $url . " (" . $enclosure->length . " Bytes) .. "; + $file = $ff->fetch( to => '/tmp' ) or die $ff->error; + print "ok\n"; + + print "summary:\n" . $sum_title . "\n\n" . $sum_text . "\n"; + + rhimport::pv_add_note($sum_title, $sum_text, $PV_ID, sprintf("%04d-%02d-%02d", $import_date[0], $import_date[1], $import_date[2]), "1"); + print "\n"; + + last; + } + } + last if $file ne ""; +} + +if($file eq "" || !(-e "$file")) { + print "No Entry found from $bd or download error - "; + if($#ARGV >= 0 && $ARGV[0] eq 'last') { + print "giving up, manual import necessary!!!\n"; + } else { + print "will retry later\n"; + } + exit 1; +} + +print "will import $file to dropbox $dropbox (cart=$to_cart)\n"; + +my $error_cb = sub { + my ($text) = @_; + + print "\n$text"; + return 0; +}; + +rhimport::check_key_file() or die "Import Key not found, use rhimport-create-id to create one\n"; + +my $ret; +my $log = rhimport::clear_carts($dbh, $group, $to_cart); +my $import_log; +($ret, $import_log) = rhimport::import_single($file, $dropbox, $user, 0, $error_cb); +$log .= $import_log; + +$dbh->disconnect(); + +unlink $file; + +if(!$ret) { + print "\nImport Error:\n"; + print $log; + exit 1; +} + +unlink($STAT_FILE); +open(my $fhs, '>', $STAT_FILE); +print $fhs $out_file; +close($fhs); + +exit 0; |